Repeated Capacitor Discharge / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

Repeated Capacitor Discharge

Move Thread

LAN_403

randommscience117

Mon Jan 02 2012, 07:35AM

Registered Member #4274 Joined: Mon Dec 19 2011, 03:10AM
Location:
Posts: 47

So for the prototype of my 3 stage 8000 J per stage coilgun project I am building 6 different coils with 6 different gauges of wire and tapping locations within the coil for different lengths of coil and different number of turns, etc. to a total of 54 combinations of coils. To test, I will charge one bank up to probably 600v (4230 Joules) and discharge it into the coil 5 times and measure the velocity each time and repeat for the 54 combinations. So, my bank of electrolytics will be discharged into 30 AWG to 12 AWG coils over 200 times, just for testing.
My problem is this, the capacitors are already used, and I don't know what repeated discharges are going to do to them. They are vented caps and in a box, so any explosions will be contained, but I don't want any of them to be ruined since they were donated and I'd have a problem getting more. So, what can I do to prevent any problems with them?
I don't want to make the bank smaller, because I don't know how well the power will scale down like that, but I'm open to suggestions. Thanks!

m4ge123

Tue Jan 03 2012, 02:47AM

Registered Member #4118 Joined: Mon Oct 03 2011, 04:50PM
Location: MD
Posts: 140

I would just make sure you have a big diode antiparallel with your bank to prevent any reverse voltage.

klugesmith

Tue Jan 03 2012, 02:50AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1714

randommscience117 wrote ...

To test, I will charge one bank up to probably 600v (4230 Joules) and discharge it into the coil 5 times and measure the velocity each time and repeat for the 54 combinations.

I sincerely applaud the brute-force, methodical, experimental approach.
But have you forgotten some very important independent variables?
For example:
The position of the projectile at the beginning of each current pulse?
Initial velocity of projectile (for second and third stage coils) ?
Coil resistance increase due to heating, if not given time to cool after each multi-kJ discharge?

Some effort at modeling, based on principles well known and documented, could greatly reduce the need to exhaustively explore your design space.

What is your most energetic previously-built-and-measured coilgun?

[edit] Re. your direct question in OP, about abuse of your capacitors. Have you estimated the magnitude and duration of your current pulses, per capacitor? Looked for any relevant OEM literature published by the maker of your capacitors, or similar ones?

Not to disrespect any "rules of thumb" from more experienced coilgunners.

[edit edit] Are you serious about 30 AWG wire? I would think that if you used enough wire mass to handle even one multi-kJ discharge, the coil inductance and resulting time constant would be impractically large. Are you familiar with Barry's online inductance calculator and RLC circuit simulator?

randommscience117

Tue Jan 03 2012, 05:01AM

Registered Member #4274 Joined: Mon Dec 19 2011, 03:10AM
Location:
Posts: 47

Well what i'm doing is going about and just testing coils with all constants such as the projectile position, and waiting for the coil to cool down, etc. I am mainly interested in seeing if a longer coil is better or worse than a shorter coil, and if more windings is better or worse. All of that through experimental data vs. Barry's coilgun simulator. As for the multistages, this again is just to test the coil shape, and once the coilgun nears completion, I will progressively add stages.
I did look up the datasheets for all three types of capacitors that will be used in the final design (there are 4700mF, 5600mF, and 6800mF caps) and one mentioned endurance testing of the capacitor. It said that the capacitor was charged and discharged 20 million times and remained within the Â±20% capacitance of the capacitor, so I guess repeated discharge isn't all that bad, of course that probably wasn't direct short circuiting, but rather a slow discharge as would be used in the inverter/audio equipment they were meant for.
This is actually my first coilgun, so small stuff here.
I have been looking into measuring the current, but I don't know exactly how coilgunners have done it. All I've seen is an oscilloscope set up and it records/displays the current readings in curves. Which reminds me of what I did see on Barry's coilgun sight. He said that the waveform seen in the oscilloscope should be a gentle curve to zero with no reverse voltage that would damage the capacitors. He had a way to calculate it using inductance and resistance, but I'm not sure exactly how to find that without having a prebuilt coil to measure those values. Also, I'm not sure of the exact power of the bank, since my multimeter measures resistance and voltage, and nothing else, so I don't know if the capacitors are what they say they are.
I want the coil to have zero reverse voltage without any unnecessary diodes or whatever. I need more accuracy. The other thing is my coils will be cooled by liquid nitrogen, so the only way I can really test the resistance precisely is by making a coil, measuring the resistance, then submerging it in liquid nitrogen to measure the resistance again and change the values if necessary.
As far as I know, this will be the most powerful coilgun on the internet ever built, so nobody really has gone through all the careful planning for something this big. I imagine i'll be needing a better multimeter soon.

klugesmith

Tue Jan 03 2012, 05:36AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1714

Thanks for stating your goal up front.
Good luck!

A few responses to individual points.

You are not going to directly short-circuit your capacitors -- not even close.
You are connecting them to a reasonably well-known load, which will
determine the discharge current and time. The load must have a much
higher impedance than the capacitor's ESR and ESL, if you expect noteworthy efficiency.

You can calculate your coils' resistance at room temperature and at LN2 temperature, with sufficient accuracy, from the wire gauge and total length (or weight). Or measure the resistance at room temperature and look up the reduction factor for 77K (about 1/7 IIRC).
That'll give you the resistance at the -beginning- of the heating pulse. Might as well assume adiabatic heating, with 100% of the capacitor's energy, to get the temperature and resistance at the -end- of the pulse.

For coilgun frequencies (100s of Hz) you can measure current pulses with an oscilloscope and current transformer, or Rogowski coil, or plain old current shunt.

You might enjoy this animation and abstract of some LN2-cooled 60- and 30-tesla pulsed magnets:

Have you read Barry's discussion of critical damping, which is the simplest solution to your objective of no voltage reversal or clamp diodes? IIRC, the peak magnetic field energy is 1/(e^2) times the initial capacitor energy, vs. 100% in an undamped LC circuit. This forum recently had a discussion about better efficiency using an underdamped RLC, with the main SCR switched off before the zero voltage point.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob